Scribd
Upload
114 views3 pages

1D FDTD Solution For Mur

This document describes a 1D finite-difference time-domain (FDTD) simulation that solves Maxwell's equations using Mur's absorbing boundary condition. The simulation models an x-directed, z-polarized transverse electromagnetic wave containing Hy and Ez fields on a grid defined by Yee's algorithm. Fields are updated at each time step using leapfrog time-stepping. Four different source types can be modeled: unit step, impulse, Gaussian, and sinusoidal. Field values at the boundaries are calculated using Engquist-Majda one-way wave equations to absorb incoming waves.

Uploaded by

Bahar
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
114 views3 pages

1D FDTD Solution For Mur

This document describes a 1D finite-difference time-domain (FDTD) simulation that solves Maxwell's equations using Mur's absorbing boundary condition. The simulation models an x-directed, z-polarized transverse electromagnetic wave containing Hy and Ez fields on a grid defined by Yee's algorithm. Fields are updated at each time step using leapfrog time-stepping. Four different source types can be modeled: unit step, impulse, Gaussian, and sinusoidal. Field values at the boundaries are calculated using Engquist-Majda one-way wave equations to absorb incoming waves.

Uploaded by

Bahar
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 3

% "1D FDTD solution for Mur's Absorbing Boundary Condition (ABC)"

%
% Objective of the program is to solve for the Maxwell's equation for an
% x-directed z-polarized TEM wave containg the y-directed magnetic
% field Hy and z-directed electric field Ez. The fields are updated at every
% timestep, in a space, where all physical parameters of free space are
% not normalized to 1 (given real and known values), using standard update
% equations obtained from the difference form of Maxwell's curl equations
% with very low electric and magnetic conductivities of 4 x10^(-4) units
% incorporated in them. The field points are defined in a grid described
% by Yee's algorithm. The H fields are defined at every half coordinate of
% a spacestep and E field is defined at every coordinate point. Also,
% the time update is done using Leapfrog time-stepping. Here, the H fields
% are updated every half time-step and E fields are updated every full
% time-step. This is shown by two alternating vector updates spanning only
% a part of spatial grid where the wave, starting from source, has reached
% at that particular time instant avoiding field updates at all points in
% grid which is unnecessary at that time instant. These spatial updates are
% inside the main for-loop for time update, spanning the entire time grid.
% Also here, the space-step length is taken as 1 micron instead of 1 unit
% in unitless domain assumed in previous programs. Also, here, the vectors
% used as multiplication factors for update equations are initialized
% before the loop starts to avoid repeated calculation of the same in every
% loop iteration, a minor attempt at optimization. The boundary condition
% here is Mur's Absorbing Boundary Condition (ABC) where the fields at the
% grid points have electric field values formulated using Engquist Majda one
% way wave equations [1] where the boundaries give a sense of absorbing the
% total field incident on them and reflecting none back to the domain.
%
% A source of electric field is defined at the center of the spatial domain
% which is a hard source, in that it does not change its value due to
% interference from external fields i.e in other words, the source is a
% perfect electric conductor. The form of the source can be varied using
% the variables sine, gaussian and impulse. The source is available in four
% standard forms- Unit-time step, Impulse, Gausian and Sinusoidal forms.
% For the sinusoidal source, the user is allowed to give a frequency of
% his/her choice in Hz. The plot of Ez field v/s spacesteps is shown at
% every time step. The simulation can be ended by closing this plot
% window or by waiting till all the time step updates are completed.
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
%Clearing variables in memory and Matlab command screen
clear all;
clc;
% Grid Dimension in x (xdim) direction
xdim=200;
%Total no of time steps
time_tot=350;
%Position of the source (center of the domain)
xsource=100;
%Courant stability factor
S=1;
% Parameters of free space (permittivity and permeability and speed of
% light) are all not 1 and are given real values
epsilon0=(1/(36*pi))*1e-9;
mu0=4*pi*1e-7;
c=3e+8;
% Spatial grid step length (spatial grid step = 1 micron)
delta=1e-6;
% Temporal gris step obtained using Courant condition
deltat=S*delta/c;
% Initialization of field vectors
Ez=zeros(1,xdim);
Hy=zeros(1,xdim);
% Initialization of permittivity and permeability vectors
epsilon=epsilon0*ones(1,xdim);
mu=mu0*ones(1,xdim);
% Initializing electric and magnetic field vectors
sigma=4e-4*ones(1,xdim);
sigma_star=4e-4*ones(1,xdim);
%Choice of natue of source
gaussian=0;
sine=0;
% The user can give a frequency of his choice for sinusoidal (if sine=1
above) waves in Hz
frequency=1.5e+13;
impulse=0;
%Choose any one as 1 and rest as 0. Default (when all are 0): Unit time step
%Multiplication factor vectors for H vector update to avoid being calculated
many times
%in the time update loop so as to increase computation speed
A=((mu-0.5*deltat*sigma_star)./(mu+0.5*deltat*sigma_star));
B=(deltat/delta)./(mu+0.5*deltat*sigma_star);

%Multiplication factor vectors for E vector update to avoid being calculated


many times
%in the time update loop so as to increase computation speed
C=((epsilon-0.5*deltat*sigma)./(epsilon+0.5*deltat*sigma));
D=(deltat/delta)./(epsilon+0.5*deltat*sigma);

% Update loop begins


for n=1:1:time_tot

%if source is impulse or unit-time step


if gaussian==0 && sine==0 && n==1
Ez(xsource)=1;
end

% Setting time dependent boundaries to update only relevant parts of the


% vector where the wave has reached to avoid unnecessary updates.
if n<xsource-2
n1=xsource-n-1;
else
n1=1;
end
if n<xdim-1-xsource
n2=xsource+n;
else
n2=xdim-1;
end

%Vector Update instead of for loop for Hy field incorporating magnetic


%conductivity
Hy(n1:n2)=A(n1:n2).*Hy(n1:n2)+B(n1:n2).*(Ez(n1+1:n2+1)-Ez(n1:n2));
%Vector Update instead of for loop for Ez field incorporating magnetic
%conductivity
Ez(n1+1:n2+1)=C(n1+1:n2+1).*Ez(n1+1:n2+1)+D(n1+1:n2+1).*(Hy(n1+1:n2+1)-
Hy(n1:n2));

% Source conditions
if impulse==0
% If unit-time step
if gaussian==0 && sine==0
Ez(xsource)=1;
end
%if sine
if sine==1
tstart=1;
N_lambda=c/(frequency*delta);
Ez(xsource)=sin(((2*pi*(c/(delta*N_lambda))*(n-tstart)*deltat)));
end
%if gaussian
if gaussian==1
if n<=42
Ez(xsource)=(10-15*cos(n*pi/20)+6*cos(2*n*pi/20)-
cos(3*n*pi/20))/32;
end
end
else
%if impulse
Ez(xsource)=0;
end

% Mur's Absorbing boundary condition


if n>1
Ez(xdim)=dum2+((S-1)/(S+1))*(Ez(xdim-1)-Ez(xdim));
Ez(2)=dum1+((S-1)/(S+1))*(Ez(2)-Ez(1));
Ez(1)=Ez(2);
end
dum1=Ez(3);
dum2=Ez(xdim-1);

plot((1:1:xdim)*delta,Ez,'color','k');
titlestring=['\fontsize{20}Plot of Ez vs x for 1D FDTD Mur Absorbing
Boundary Condition at time = ',num2str(round(n*deltat/10e-15)),' fs'];
title(titlestring,'color','k');
xlabel('x in m','FontSize',20);
ylabel('Ez in V/m','FontSize',20);
set(gca,'FontSize',20);
axis([0 xdim*delta -3 3]);
getframe;
end

You might also like